Combining data from segregation analyses and mutation screening studies, the established breast cancer susceptibility genes are responsible for an estimated 20%-25% of the genetic component of this disease. The genes and/or sequence variants responsible for the remaining genetic component of breast cancer risk have yet to be identified. Most of the current enthusiasm for SNPs and haplotype mapping are predicated on the assumption that common modest risk variants are most important. However, few candidate associations between common SNPs and breast cancer risk have been independently reproduced. Thus the central question of this study: What is the relative contribution of common (usually modest-risk) sequence variants vs. rare (potentially higher-risk) sequence variants to the genetic attributable fraction of breast cancer? U Using an ethnically diverse series of 1,250 genetically high-risk breast cancer cases and 1,250 frequency-matched population controls, we propose a novel study designed to make a direct comparison between the common disease/ common variant and common disease/ rare variant models of genetic susceptibility. The study has two arms. In the first, we will genotype the cases and controls with all of the common- sequence variants that are known, or are found over the course of this study by the breast cancer genetics research community, to predict increased risk of breast cancer. In the second arm, we will mutation screen the open reading frames of strong candidate susceptibility genes in both the cases and the controls. Analysis of the genotype and mutation screening data should provide an answer to the central study question. Our focus on early onset and familial cases will substantially increase power to detect risk conferred by deleterious sequence variants as compared to a study of similar size without these criteria, fl Results from this study are relevant to public health in three ways: (1) This study will provide a hypothesis test of genes, and mutations in them, that appear to confer moderately to dramatically increased risk of breast cancer. Measuring risk due to mutations in these genes is a key step that lies between initial indications that the gene plays a role in breast cancer susceptibility and bringing the gene into the clinical practice of cancer genetics. (2) Results from this study will bear on the future direction of clinical cancer genetics. The relative contribution that moderate risk versus modest risk sequence variants make to the attributable risk of breast cancer will have an impact on how the genetic information enters clinical practice. (3) Analysis of the genotype and mutation screening data will provide a comparison of risk attributable to the common variant and rare variant genetic models of cancer susceptibility. This is a question of major current interest and importance within the genetics research community. If we observe that the rare sequence variants account for as much or more risk than do common SNPs, it may be necessary to expand mutation screening from the realm of genetic epidemiology/ family studies into larger scale population-based studies.